exceptions_exp2_swap_0.3_last_to_carry_40817

This model is a fine-tuned version of on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 3.5438
• Accuracy: 0.3747

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.0006
• train_batch_size: 16
• eval_batch_size: 16
• seed: 40817
• gradient_accumulation_steps: 5
• total_train_batch_size: 80
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.98) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 100
• num_epochs: 50.0
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Accuracy	Validation Loss
4.8306	0.2915	1000	0.2526	4.7673
4.3469	0.5830	2000	0.2979	4.2918
4.1502	0.8745	3000	0.3138	4.1034
3.9947	1.1659	4000	0.3238	3.9996
3.9426	1.4574	5000	0.3306	3.9209
3.8974	1.7488	6000	0.3363	3.8623
3.7568	2.0402	7000	0.3404	3.8194
3.7566	2.3317	8000	0.3439	3.7866
3.7595	2.6232	9000	0.3462	3.7594
3.7283	2.9147	10000	0.3490	3.7324
3.6411	3.2061	11000	0.3506	3.7193
3.6554	3.4976	12000	0.3522	3.7021
3.6491	3.7891	13000	0.3543	3.6857
3.5521	4.0805	14000	0.3552	3.6766
3.5956	4.3719	15000	0.3563	3.6657
3.5821	4.6634	16000	0.3574	3.6508
3.593	4.9549	17000	0.3589	3.6380
3.5122	5.2463	18000	0.3592	3.6424
3.5202	5.5378	19000	0.3600	3.6308
3.5335	5.8293	20000	0.3610	3.6190
3.4488	6.1207	21000	0.3614	3.6236
3.4823	6.4122	22000	0.3618	3.6163
3.4926	6.7037	23000	0.3629	3.6068
3.5005	6.9952	24000	0.3635	3.5961
3.4259	7.2865	25000	0.3637	3.6062
3.4517	7.5780	26000	0.3645	3.5967
3.4629	7.8695	27000	0.3649	3.5859
3.3778	8.1609	28000	0.3649	3.5964
3.4225	8.4524	29000	0.3653	3.5901
3.4337	8.7439	30000	0.3664	3.5808
3.336	9.0353	31000	0.3660	3.5837
3.3834	9.3268	32000	0.3664	3.5845
3.4056	9.6183	33000	0.3670	3.5777
3.4196	9.9098	34000	0.3674	3.5701
3.3499	10.2011	35000	0.3669	3.5815
3.3791	10.4926	36000	0.3677	3.5725
3.3851	10.7841	37000	0.3680	3.5672
3.3093	11.0755	38000	0.3681	3.5752
3.3482	11.3670	39000	0.3684	3.5698
3.3554	11.6585	40000	0.3686	3.5652
3.3859	11.9500	41000	0.3693	3.5546
3.2994	12.2414	42000	0.3688	3.5683
3.3338	12.5329	43000	0.3692	3.5629
3.3632	12.8243	44000	0.3700	3.5532
3.2733	13.1157	45000	0.3692	3.5695
3.3182	13.4072	46000	0.3698	3.5623
3.3392	13.6987	47000	0.3702	3.5540
3.3438	13.9902	48000	0.3710	3.5478
3.2865	14.2816	49000	0.3701	3.5607
3.3122	14.5731	50000	0.3704	3.5550
3.3241	14.8646	51000	0.3711	3.5480
3.2389	15.1559	52000	0.3704	3.5644
3.2726	15.4474	53000	0.3706	3.5572
3.3046	15.7389	54000	0.3713	3.5486
3.2146	16.0303	55000	0.3706	3.5583
3.2738	16.3218	56000	0.3712	3.5576
3.2802	16.6133	57000	0.3712	3.5505
3.3029	16.9048	58000	0.3718	3.5417
3.2261	17.1962	59000	0.3709	3.5586
3.2492	17.4877	60000	0.3713	3.5544
3.2874	17.7792	61000	0.3720	3.5445
3.2104	18.0705	62000	0.3714	3.5553
3.2364	18.3620	63000	0.3719	3.5509
3.2566	18.6535	64000	0.3720	3.5489
3.2744	18.9450	65000	0.3726	3.5388
3.203	19.2364	66000	0.3719	3.5590
3.2497	19.5279	67000	0.3723	3.5491
3.2628	19.8194	68000	0.3725	3.5406
3.1847	20.1108	69000	0.3721	3.5564
3.2097	20.4023	70000	0.3719	3.5550
3.2618	20.6938	71000	0.3727	3.5426
3.2497	20.9853	72000	0.3730	3.5406
3.2048	21.2766	73000	0.3721	3.5574
3.2305	21.5681	74000	0.3729	3.5447
3.2455	21.8596	75000	0.3734	3.5377
3.1695	22.1510	76000	0.3725	3.5593
3.2005	22.4425	77000	0.3731	3.5495
3.2406	22.7340	78000	0.3733	3.5460
3.1306	23.0254	79000	0.3727	3.5534
3.1794	23.3169	80000	0.3730	3.5510
3.1802	23.6083	81000	3.5560	0.3726
3.2111	23.8998	82000	3.5481	0.3731
3.1603	24.1915	83000	3.5617	0.3722
3.1829	24.4830	84000	3.5501	0.3734
3.2145	24.7745	85000	3.5431	0.3736
3.1312	25.0659	86000	3.5575	0.3729
3.1629	25.3574	87000	3.5505	0.3734
3.1895	25.6489	88000	3.5468	0.3735
3.2135	25.9404	89000	3.5367	0.3739
3.152	26.2317	90000	3.5540	0.3732
3.1776	26.5232	91000	3.5495	0.3736
3.194	26.8147	92000	3.5429	0.3739
3.1246	27.1061	93000	3.5594	0.3728
3.1573	27.3976	94000	3.5525	0.3734
3.1761	27.6891	95000	3.5472	0.3738
3.1895	27.9806	96000	3.5391	0.3743
3.144	28.2720	97000	3.5557	0.3735
3.1599	28.5635	98000	3.5458	0.3742
3.1835	28.8550	99000	3.5436	0.3741
3.1266	29.1463	100000	3.5591	0.3735
3.1389	29.4378	101000	3.5503	0.3739
3.1715	29.7293	102000	3.5423	0.3743
3.0879	30.0207	103000	3.5552	0.3737
3.1272	30.3122	104000	3.5557	0.3737
3.1359	30.6037	105000	3.5508	0.3740
3.1669	30.8952	106000	3.5402	0.3747
3.1063	31.1866	107000	3.5585	0.3741
3.1367	31.4781	108000	3.5508	0.3743
3.1412	31.7695	109000	3.5438	0.3747

Framework versions

• Transformers 4.55.2
• Pytorch 2.8.0+cu128
• Datasets 4.0.0
• Tokenizers 0.21.4